Dynamic State Estimation of Power System Based on Improved Adaptive Unscented Particle Filter

In this paper, an improved adaptive unscented particle filter (IAUPF) algorithm is proposed to address the shortcomings of the unscented particle filter (UPF) algorithm on the state estimation of distribution network, such as vulnerability to process noise and low quality of the importance density function, in order to obtain a more accurate state estimation result and reduce the effect of unknown system noise in the dynamic state estimation. The IAUPF can estimate the mean and variance of the system noise and so increase the filtering accuracy of the system with unknown noise by employing a novel statistical estimator for the noise parameter and modifying the scale correction factor in real-time. The simulation results on the IEEE 33-node system show that, as opposed to the conventional UPF algorithm, the proposed IAUPF can address the issue of decreasing estimation accuracy due to unknown system noise in the filtering process and ensure high precision of state estimation when the system experiences abrupt changes.